Many accessibility features come with an unspoken requirement: the person who needs them must also know how to find, configure and maintain them.

You need the right device, the right application and the right combination of settings. When something stops working, the user who already faces the barrier is often expected to diagnose the problem.

That is why a recent trial by O2 and Mavenir caught my attention. It explored what happens when some of the accessibility work moves away from the individual device and into the network itself.

The result points to something much bigger than clearer telephone calls.

Moving accessibility into the network

O2 and network software provider Mavenir ran a proof-of-concept designed to improve calls for people with hearing loss.

Participants completed a short, automated hearing test that identified how they perceived different sound frequencies. This created a personalised hearing profile, which was securely linked to their mobile number.

During a call, the network used that profile to adjust how speech was delivered in real time. Participants could continue using their existing phone and telephone number. They did not need to install another application or change how they normally made a call.

According to Virgin Media O2, nearly 90 per cent of participants reported improved call clarity. They also described reduced listening effort, fewer misunderstandings and a more natural experience.

Mary Higgins, who is profoundly deaf and took part in the trial, described the practical difference:

“I could hear clearly without my hearing aids and didn’t need to keep asking people to repeat themselves.”

That is a short quote, but it captures the outcome that matters. She could participate in the conversation with less effort.

These are encouraging findings, although they need to be kept in perspective. This was an early proof-of-concept, not the launch of a commercial service. The announcement does not provide a sample size or detailed methodology, so I see the result as promising rather than conclusive.

Even so, the direction is important.

The best feature may be the one you never switch on

I use assistive technology every day. I know its value, but I also know how easily the burden of accessibility can fall back onto the user.

You are expected to find the appropriate setting, understand what it does and work out why it conflicts with something else. You may then have to repeat the process when you replace a device, update some software or move to another service.

O2’s trial takes a different approach. Instead of asking what extra product a person with hearing loss should acquire, it asks what the underlying service can do differently.

That shift matters.

Accessibility is often discussed at the interface level. We look at buttons, colour contrast, keyboard operation, captions and compatibility with assistive technologies. These things are essential, but the visible interface is only one part of the experience.

The network, platform, data model and shared services beneath it can either remove barriers or quietly create them.

A well-designed interface cannot fully compensate for an inaccessible system underneath it.

Designed for some, useful to many

The O2 trial was created to address a particular barrier experienced by people with hearing loss. We should not dilute that purpose or pretend that its current implementation automatically benefits every customer.

However, accessibility-led innovation has a long history of producing wider benefits.

Captions support Deaf and hard-of-hearing people, but they are also used in busy stations, quiet offices and homes where somebody is trying not to wake a sleeping child. Voice control can be essential for someone who cannot use a touchscreen, while also being convenient for someone cooking with both hands occupied.

Clear language helps people with cognitive impairments. It also helps anyone reading under pressure or working in a second language.

The wider lesson is not that one solution will work for everybody. It is that systems designed to accommodate human variation tend to be more flexible than systems built around an imaginary average person.

For telephone calls, adaptable audio might eventually support people experiencing temporary hearing changes, fatigue or difficult listening conditions. The important innovation is not one perfect audio profile.

It is a network capable of responding to different needs.

Ageing makes this a mainstream design issue

The need for this thinking will only grow.

The World Health Organization says that more than a quarter of people aged over 60 are affected by disabling hearing loss. It also projects that nearly 2.5 billion people worldwide will have some degree of hearing loss by 2050.

The UK population is ageing too. The Office for National Statistics has projected that approximately one in four people in the UK will be aged 65 or over by 2050.

This is not a niche audience waiting somewhere in the future. It includes our customers, employees, colleagues, friends and family members. Eventually, it may include any one of us.

Ageing can bring changes in hearing, sight, dexterity, memory and stamina. These changes do not remove a person’s need to communicate, work, manage money, obtain healthcare or participate in their community.

If essential services remain usable only by people with a narrow range of abilities, more people will find themselves excluded as the population ages.

Accessible infrastructure can help people retain something extremely important: independence.

A clearer call can mean speaking to a doctor without relying on somebody else to interpret. An accessible banking service can mean managing your own finances. A usable transport application can mean travelling without asking another person to complete part of the journey.

These may look like small interactions to a service designer. Taken together, they determine how independently someone can live.

There is also a straightforward commercial point. An ageing population means a growing number of customers who will value services that remain easy to use as their needs change.

Organisations can design for those customers now, or spend considerably more trying to retrofit inaccessible systems later.

Invisible must not mean uncontrollable

There is another side to background personalisation.

A hearing profile is sensitive personal information. If it is linked to a telephone number and used automatically, customers need to understand what is being collected, where it is stored and who can access it.

They should be able to inspect, change and remove the profile. Participation should be based on meaningful consent, not buried in a long set of terms. The service should also be tested with people who have different types and degrees of hearing loss.

There are practical questions too. Would the optimisation continue when roaming? What happens when someone changes network? Could it interact unexpectedly with hearing aids, headphones or audio settings on the device?

These are not reasons to reject the idea. They are requirements to address if it moves from a trial into a real service.

Seamless accessibility should reduce effort without removing agency.

That principle will become increasingly important as organisations use artificial intelligence and automated personalisation to shape customer experiences.

Design the burden out

O2’s experiment gives organisations a useful question to ask:

What accessibility work are we currently making the user do?

Look beyond the visible interface. Consider the complete service, including its infrastructure, identity systems, communications, customer support and data.

Then ask:

1. Can the underlying system remove or reduce the barrier?
2. Can personalisation happen without requiring specialist technical knowledge?
3. Does the user retain meaningful choice and control?
4. Have disabled people been involved in designing and testing it?
5. Are we measuring independence and successful outcomes, rather than simply whether the feature operates?

Accessibility should not mean hiding every setting or making assumptions on someone’s behalf. Nor should it mean replacing specialist assistive technology that people already rely on.

It means building services that recognise variation from the beginning and do more of the work themselves.

Accessibility at the heart of the service

What makes the O2 trial interesting is not simply the possibility of clearer calls for people with hearing loss, valuable though that could be.

It demonstrates what becomes possible when accessibility moves from the edge of a service into the infrastructure at its heart.

As our population ages, more people will encounter changes in how they see, hear, remember and interact. Designing for those changes is not an act of charity or a specialist exercise. It is how we keep customers included, connected and independent.

Build around a narrow idea of the average user and more people will be excluded over time.

Build for human variation, and everybody gets a better chance to participate.

Sources

• O2 and Mavenir successfully trial technology to improve call clarity for people with hearing loss, Virgin Media O2, 11 June 2026
• Deafness and hearing loss, World Health Organization
• Overview of the UK population, Office for National Statistics